From: Roland McGrath <roland@redhat.com>
Subject: [RHEL5 PATCH] fix vdso in core dumps (RHBZ #211744)
Date: Sat, 13 Jan 2007 21:49:44 -0800 (PST)
Bugzilla: 211744
Message-Id: <20070114054944.D52A61800E5@magilla.sf.frob.com>
Changelog: fix vdso in core dumps


The problems are inherited from upstream.  There used to be corrections for
this incorporated into the exec-shield patch when it had a lot of change to
the vdso setup; it no longer has to do anything to the vdso code because
upstream took similar changes, but broke the core dump part.  Note that
linux-2.6-execshield.patch has some related cruft from the past that should
not be there, and I've excised it from the rawhide version.

Thanks,
Roland


--- linux-2.6.18/arch/i386/kernel/sysenter.c
+++ linux-2.6.18/arch/i386/kernel/sysenter.c
@@ -93,15 +93,6 @@
 #ifdef CONFIG_COMPAT_VDSO
 	__set_fixmap(FIX_VDSO, __pa(syscall_page), PAGE_READONLY);
 	printk("Compat vDSO mapped to %08lx.\n", __fix_to_virt(FIX_VDSO));
-#else
-	/*
-	 * In the non-compat case the ELF coredumping code needs the fixmap:
-	 */
-#ifdef CONFIG_XEN
-	__set_fixmap(FIX_VDSO, virt_to_machine(syscall_page), PAGE_KERNEL_RO);
-#else
-	__set_fixmap(FIX_VDSO, __pa(syscall_page), PAGE_KERNEL_RO);
-#endif
 #endif
 
 	if (!boot_cpu_has(X86_FEATURE_SEP)) {
@@ -165,6 +156,13 @@
 	vma->vm_end = addr + PAGE_SIZE;
 	/* MAYWRITE to allow gdb to COW and set breakpoints */
 	vma->vm_flags = VM_READ|VM_EXEC|VM_MAYREAD|VM_MAYEXEC|VM_MAYWRITE;
+	/*
+	 * Make sure the vDSO gets into every core dump.
+	 * Dumping its contents makes post-mortem fully interpretable later
+	 * without matching up the same kernel and hardware config to see
+	 * what PC values meant.
+	 */
+	vma->vm_flags |= VM_ALWAYSDUMP;
 	vma->vm_flags |= mm->def_flags;
 	vma->vm_page_prot = protection_map[vma->vm_flags & 7];
 	vma->vm_ops = &syscall_vm_ops;
--- linux-2.6.18/arch/powerpc/kernel/vdso.c
+++ linux-2.6.18/arch/powerpc/kernel/vdso.c
@@ -283,6 +283,13 @@
 	 * pages though
 	 */
 	vma->vm_flags = VM_READ|VM_EXEC|VM_MAYREAD|VM_MAYWRITE|VM_MAYEXEC;
+	/*
+	 * Make sure the vDSO gets into every core dump.
+	 * Dumping its contents makes post-mortem fully interpretable later
+	 * without matching up the same kernel and hardware config to see
+	 * what PC values meant.
+	 */
+	vma->vm_flags |= VM_ALWAYSDUMP;
 	vma->vm_flags |= mm->def_flags;
 	vma->vm_page_prot = protection_map[vma->vm_flags & 0x7];
 	vma->vm_ops = &vdso_vmops;
--- linux-2.6.18/arch/x86_64/ia32/ia32_binfmt.c
+++ linux-2.6.18/arch/x86_64/ia32/ia32_binfmt.c
@@ -64,55 +64,6 @@
 #define ELF_NGREG (sizeof (struct user_regs_struct32) / sizeof(elf_greg_t))
 typedef elf_greg_t elf_gregset_t[ELF_NGREG];
 
-/*
- * These macros parameterize elf_core_dump in fs/binfmt_elf.c to write out
- * extra segments containing the vsyscall DSO contents.  Dumping its
- * contents makes post-mortem fully interpretable later without matching up
- * the same kernel and hardware config to see what PC values meant.
- * Dumping its extra ELF program headers includes all the other information
- * a debugger needs to easily find how the vsyscall DSO was being used.
- */
-#define ELF_CORE_EXTRA_PHDRS	(find_vma(current->mm, VSYSCALL32_BASE) ?     \
-    (VSYSCALL32_EHDR->e_phnum) : 0)
-#define ELF_CORE_WRITE_EXTRA_PHDRS					      \
-do {									      \
-	if (find_vma(current->mm, VSYSCALL32_BASE)) { 			      \
-		const struct elf32_phdr *const vsyscall_phdrs =		      \
-			(const struct elf32_phdr *) (VSYSCALL32_BASE	      \
-						   + VSYSCALL32_EHDR->e_phoff);\
-		int i;							      \
-		Elf32_Off ofs = 0;					      \
-		for (i = 0; i < VSYSCALL32_EHDR->e_phnum; ++i) {	      \
-			struct elf32_phdr phdr = vsyscall_phdrs[i];	      \
-			if (phdr.p_type == PT_LOAD) {			      \
-				BUG_ON(ofs != 0);			      \
-				ofs = phdr.p_offset = offset;		      \
-				phdr.p_memsz = PAGE_ALIGN(phdr.p_memsz);      \
-				phdr.p_filesz = phdr.p_memsz;		      \
-				offset += phdr.p_filesz;		      \
-			}						      \
-			else						      \
-				phdr.p_offset += ofs;			      \
-			phdr.p_paddr = 0; /* match other core phdrs */	      \
-			DUMP_WRITE(&phdr, sizeof(phdr));		      \
-		}							      \
-	}								      \
-} while (0)
-#define ELF_CORE_WRITE_EXTRA_DATA					      \
-do {									      \
-	if (find_vma(current->mm, VSYSCALL32_BASE)) { 			      \
-		const struct elf32_phdr *const vsyscall_phdrs =		      \
-			(const struct elf32_phdr *) (VSYSCALL32_BASE	      \
-						   + VSYSCALL32_EHDR->e_phoff);      \
-		int i;							      \
-		for (i = 0; i < VSYSCALL32_EHDR->e_phnum; ++i) {	      \
-			if (vsyscall_phdrs[i].p_type == PT_LOAD)	      \
-				DUMP_WRITE((void *) (u64) vsyscall_phdrs[i].p_vaddr,\
-				    PAGE_ALIGN(vsyscall_phdrs[i].p_memsz));   \
-		}							      \
-	}								      \
-} while (0)
-
 struct elf_siginfo
 {
 	int	si_signo;			/* signal number */
--- linux-2.6.18/arch/x86_64/ia32/syscall32.c
+++ linux-2.6.18/arch/x86_64/ia32/syscall32.c
@@ -61,6 +61,13 @@
 	vma->vm_end = VSYSCALL32_END;
 	/* MAYWRITE to allow gdb to COW and set breakpoints */
 	vma->vm_flags = VM_READ|VM_EXEC|VM_MAYREAD|VM_MAYEXEC|VM_MAYWRITE;
+	/*
+	 * Make sure the vDSO gets into every core dump.
+	 * Dumping its contents makes post-mortem fully interpretable later
+	 * without matching up the same kernel and hardware config to see
+	 * what PC values meant.
+	 */
+	vma->vm_flags |= VM_ALWAYSDUMP;
 	vma->vm_flags |= mm->def_flags;
 	vma->vm_page_prot = protection_map[vma->vm_flags & 7];
 	vma->vm_ops = &syscall32_vm_ops;
--- linux-2.6.18/fs/binfmt_elf.c
+++ linux-2.6.18/fs/binfmt_elf.c
@@ -1238,13 +1238,14 @@
  */
 static int maydump(struct vm_area_struct *vma)
 {
+	/* The vma can be set up to tell us the answer directly.  */
+	if (vma->vm_flags & VM_ALWAYSDUMP)
+		return 1;
+
 	/* Do not dump I/O mapped devices or special mappings */
 	if (vma->vm_flags & (VM_IO | VM_RESERVED))
 		return 0;
 
-	if (vma->vm_flags & VM_DONTEXPAND) /* Kludge for vDSO.  */
-		return 1;
-
 	/* Dump shared memory only if mapped from an anonymous file. */
 	if (vma->vm_flags & VM_SHARED)
 		return vma->vm_file->f_dentry->d_inode->i_nlink == 0;
--- linux-2.6.18/include/asm-i386/elf.h
+++ linux-2.6.18/include/asm-i386/elf.h
@@ -170,50 +170,6 @@
 		NEW_AUX_ENT(AT_SYSINFO_EHDR, VDSO_COMPAT_BASE);	\
 } while (0)
 
-/*
- * These macros parameterize elf_core_dump in fs/binfmt_elf.c to write out
- * extra segments containing the vsyscall DSO contents.  Dumping its
- * contents makes post-mortem fully interpretable later without matching up
- * the same kernel and hardware config to see what PC values meant.
- * Dumping its extra ELF program headers includes all the other information
- * a debugger needs to easily find how the vsyscall DSO was being used.
- */
-#define ELF_CORE_EXTRA_PHDRS		(VDSO_HIGH_EHDR->e_phnum)
-#define ELF_CORE_WRITE_EXTRA_PHDRS					      \
-do {									      \
-	const struct elf_phdr *const vsyscall_phdrs =			      \
-		(const struct elf_phdr *) (VDSO_HIGH_BASE		      \
-					   + VDSO_HIGH_EHDR->e_phoff);    \
-	int i;								      \
-	Elf32_Off ofs = 0;						      \
-	for (i = 0; i < VDSO_HIGH_EHDR->e_phnum; ++i) {		      \
-		struct elf_phdr phdr = vsyscall_phdrs[i];		      \
-		if (phdr.p_type == PT_LOAD) {				      \
-			BUG_ON(ofs != 0);				      \
-			ofs = phdr.p_offset = offset;			      \
-			phdr.p_memsz = PAGE_ALIGN(phdr.p_memsz);	      \
-			phdr.p_filesz = phdr.p_memsz;			      \
-			offset += phdr.p_filesz;			      \
-		}							      \
-		else							      \
-			phdr.p_offset += ofs;				      \
-		phdr.p_paddr = 0; /* match other core phdrs */		      \
-		DUMP_WRITE(&phdr, sizeof(phdr));			      \
-	}								      \
-} while (0)
-#define ELF_CORE_WRITE_EXTRA_DATA					      \
-do {									      \
-	const struct elf_phdr *const vsyscall_phdrs =			      \
-		(const struct elf_phdr *) (VDSO_HIGH_BASE		      \
-					   + VDSO_HIGH_EHDR->e_phoff);    \
-	int i;								      \
-	for (i = 0; i < VDSO_HIGH_EHDR->e_phnum; ++i) {		      \
-		if (vsyscall_phdrs[i].p_type == PT_LOAD)		      \
-			DUMP_WRITE((void *) vsyscall_phdrs[i].p_vaddr,	      \
-				   PAGE_ALIGN(vsyscall_phdrs[i].p_memsz));    \
-	}								      \
-} while (0)
-
 #endif
 
 #define __HAVE_ARCH_RANDOMIZE_BRK
--- linux-2.6.18/include/linux/mm.h
+++ linux-2.6.18/include/linux/mm.h
@@ -168,6 +168,7 @@
 #ifdef CONFIG_XEN
 #define VM_FOREIGN	0x04000000	/* Has pages belonging to another VM */
 #endif
+#define VM_ALWAYSDUMP	0x08000000	/* Always include in core dumps */
 
 #ifndef VM_STACK_DEFAULT_FLAGS		/* arch can override this */
 #define VM_STACK_DEFAULT_FLAGS VM_DATA_DEFAULT_FLAGS